Delegate MaskedAssignKernel
Where-masked bit copy of one element width over prepared, broadcast operands (np.copyto's element loop):
dst[i] = src[i] wherever mask[i] != 0.
public delegate void MaskedAssignKernel(byte* src, byte* dst, byte* mask, long* srcStrides, long* dstStrides, long* maskStrides, long* shape, int ndim)
Parameters
srcbyte*The source's first element (already cast to the destination's dtype).
dstbyte*The destination's first element.
maskbyte*The mask's first byte (a bool array; any nonzero byte is true).
srcStrideslong*Source BYTE strides, one per axis, C order (the last axis is the inner run); 0 broadcasts.
dstStrideslong*Destination BYTE strides, one per axis.
maskStrideslong*Mask BYTE strides, one per axis; 0 broadcasts.
shapelong*The extent of each axis; every extent must be at least 1.
ndimintThe number of axes, at least 1 (a 0-d assignment passes one axis of extent 1).